Development of a Recognition System for the Mexican Sign Language Alphabet
Francisco Javier López-Vázquez, Héctor A. Guerrero-Osuna, Jesús Antonio Nava-Pintor, Carlos Alberto Olvera-Olvera, Germán Díaz-Flórez, Luis Fernando Luque-Vega · 2024
Effective communication is crucial for mutual understanding and remains a significant challenge for Deaf and Hard-of-Hearing individuals. This study aims to establish the basis for creating a portable communication system that can translate Mexican Sign Language (MLS) to facilitate two-way communication for Deaf and Hard-of-Hearing individuals. The device achieves a high accuracy of 98% in recognizing individual signs and features real-time audio-to-text translation capabilities. The Random Forest algorithm demonstrated a recognition accuracy of 98.97% for each letter of the Mexican alphabet. The device maintained a confidence level of 98% under optimal lighting conditions but decreased to 83% in low-light environments. It successfully recognized signs at distances up to 100 cm, with notable performance in distinguishing visually similar signs. This study has established a foundational framework for creating a portable device that allows Deaf and Hard-of-Hearing individuals to communicate effectively. It proves the device's potential as a reliable tool for bidirectional communication and instills optimism about its potential impact.